Return to BSD News archive
Path: sserve!newshost.anu.edu.au!munnari.oz.au!bunyip.cc.uq.oz.au!harbinger.cc.monash.edu.au!yeshua.marcam.com!usc!cs.utexas.edu!uunet!news.sprintlink.net!clark.net!clark.net!not-for-mail From: ack@clark.net (Eric S. Hvozda) Newsgroups: comp.os.386bsd.bugs Subject: Re: [NetBSD 0.9] patch to wd.c for IDE controller weirdnesses Date: 3 Jan 1994 19:11:30 -0500 Organization: Clark Internet Services, Inc. Lines: 40 Message-ID: <2gac7i$ma7@explorer.clark.net> References: <2ga4l7$i75@explorer.clark.net> <2ga76e$r01@unix1.cc.uop.edu> NNTP-Posting-Host: explorer.clark.net Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit In article <2ga76e$r01@unix1.cc.uop.edu>, Ken Hughes <hughes@napa.eng.uop.edu> wrote: > >Don't know of this is related, but I had a similar problem this weekend when >I tried to add a second IDE drive to my system. The system booed fine with >one drive (wd0) but adding the second caused a panic/trap/reboot. After >capturing the fleeting message on my video recorded and finding the complaint >"no disk label on wd0", I discovered that NetBSD was ignoring the CMOS >configuration for wd0 and using the parameters as (apparently) queried from >the drive. Luckily, my master drive is a Maxtor which I could jumper to >specify the "default" configuration and after doing this things were fine. Well it was labeled before I patched wd.c, and worked fine then. It's only since I've patched wd.c for the famous IDE lock problem that this developed. I was thinking or relabeling the disk, but if it can read the label on wd1, why not wd0? It's almost as if daul IDE support was zapped when I patched it. I have the two sources and I'm gonna look into this and see if I can figure this out. It's kinda annyoing to lose disk space :-) Meanwhile I'm back on the old kernel until I can figure this one out... >I found this curious since the CMOS parameters were apparently used when >running in a single disk system, just not when the second drive was present] >on the IDE chain. I have found that if you have / on wd0 and then have paritions from wd1 mounted later on, you don't even need wd1 in CMOS. However, if you have a backup / on wd1 and expect to boot from it from the prompt, it better be in CMOS or you get nifty msgs about it missing... >On another note, I started getting the "lpt0: stray interrupt 7" errors also >when the second drive was added. I get that all the time too. It's covered in the FAQ, and it's nothing to worry about. Hell, I even get it when the printer is connected to the machine and I have a device config'd for it. I chnaged the code of lpt.c so it only complains 5 times and quits whining like isa.c :-) -- Ack! Creek, not creek; Pop not soda; Car needs washed...